geofeaturekit 0.2.4

GeoFeatureKit transforms simple coordinates into powerful geospatial insights. Analyze street networks, POI diversity, and spatial patterns with professional progress tracking – no paid APIs or complex setup required.

GeoFeatureKit

GeoFeatureKit turns raw coordinates into rich, structured geospatial features – instantly.

🎯 What You Get

Input: Just latitude and longitude coordinates
Output: Comprehensive geospatial intelligence including:

• 40+ POI categories: restaurants, hospitals, subway stations, benches, toilets, and more
• Street network metrics: connectivity, total street length, segment distributions, pattern entropy
• Spatial intelligence: POI diversity indices (Shannon, Simpson) and clustering patterns

🚀 Use Cases

Domain	Application	Key Features
🤖 Machine Learning	Price prediction, exposure analysis	Rich feature vectors, contextual embeddings
📊 Research	Propensity score matching	Urban covariates, accessibility metrics
🏙️ Urban Planning	Accessibility research, zoning analysis	Spatial patterns, connectivity measures
🧠 AI/ML	Neural networks, spatial clustering	Environmental context, amenity features

🔧 Recent Updates (v0.2.4)

✅ Fixed spatial distribution bug: Mean nearest neighbor distance now correctly calculates distances to ALL other points, not just subsequent ones
✅ Improved coordinate detection: Better handling of meter vs degree coordinate systems
✅ Enhanced precision: Cleaner formatting with appropriate decimal places
✅ Fixed network metrics: Corrected dead end and intersection counting logic
✅ Robust testing: Replaced flaky tests with deterministic grid-based validation
✅ Python 3.9+ compatibility: Full support across Python versions

✨ Why GeoFeatureKit?

Advantage	Benefit
✅ Simple	Just coordinates in – structured features out
✅ Powerful	Dozens of geospatial metrics in one function call
✅ User-friendly	Optional progress bars and verbose modes
✅ Open Data	Built entirely on OSM and public geospatial libraries

🚀 Quick Start

Installation

pip install geofeaturekit

Basic Usage

from geofeaturekit import features_from_location

# Example: Analyze Times Square with progress bar
features = features_from_location({
    'latitude': 40.7580,
    'longitude': -73.9855,
    'radius_meters': 500
}, show_progress=True)

print(features)

📝 Example Output

Times Square Analysis (500m radius):

{
  "network_metrics": {
    "basic_metrics": {
      "total_nodes": 777,
      "total_street_segments": 2313,
      "total_intersections": 731,
      "total_dead_ends": 0,
      "total_street_length_meters": 80044.7
    },
    "density_metrics": {
      "intersections_per_sqkm": 930.74,
      "street_length_per_sqkm": 101.92
    },
    "connectivity_metrics": {
              "streets_to_nodes_ratio": 1.488,
        "average_connections_per_node": {
          "value": 5.954,
          "confidence_interval_95": {
            "lower": 5.837,
            "upper": 6.071
          }
        }
    },
    "street_pattern_metrics": {
      "street_segment_length_distribution": {
        "minimum_meters": 0.5,
        "maximum_meters": 286.6,
        "mean_meters": 34.6,
        "median_meters": 12.0,
        "std_dev_meters": 50.7
      },
      "street_bearing_distribution": {
        "mean_degrees": 163.3,
        "std_dev_degrees": 101.5
      },
      "ninety_degree_intersection_ratio": 0.0,
      "bearing_entropy": 2.056
    }
  },
  "poi_metrics": {
    "absolute_counts": {
      "total_points_of_interest": 1076,
      "counts_by_category": {
        "total_restaurant_places": {
          "count": 173,
          "percentage": 16.1
        },
        "total_fast_food_places": {
          "count": 77,
          "percentage": 7.2
        },
        "total_cafe_places": {
          "count": 74,
          "percentage": 6.9
        },
        "total_bicycle_parking_places": {
          "count": 71,
          "percentage": 6.6
        },
        "total_bench_places": {
          "count": 27,
          "percentage": 2.5
        },
        "total_bar_places": {
          "count": 26,
          "percentage": 2.4
        },
        "total_bank_places": {
          "count": 24,
          "percentage": 2.2
        },
        "total_pub_places": {
          "count": 19,
          "percentage": 1.8
        },
        "total_bicycle_rental_places": {
          "count": 15,
          "percentage": 1.4
        },
        "total_theatre_places": {
          "count": 12,
          "percentage": 1.1
        },
        "total_pharmacy_places": {
          "count": 6,
          "percentage": 0.6
        },
        "total_atm_places": {
          "count": 4,
          "percentage": 0.4
        }
      }
    },
    "density_metrics": {
      "points_of_interest_per_sqkm": 1370.700637,
      "density_by_category": {
        "restaurant_places_per_sqkm": 220.382166,
        "fast_food_places_per_sqkm": 98.089172,
        "cafe_places_per_sqkm": 94.267516,
        "bicycle_parking_places_per_sqkm": 90.44586,
        "bank_places_per_sqkm": 30.573248,
        "theatre_places_per_sqkm": 15.286624,
        "pharmacy_places_per_sqkm": 7.643312
      }
    },
    "distribution_metrics": {
      "unique_category_count": 42,
              "largest_category": {
          "name": "restaurant",
          "count": 173,
          "percentage": 16.08
        },
      "diversity_metrics": {
        "shannon_diversity_index": 2.245,
        "simpson_diversity_index": 0.79,
        "category_evenness": 0.601
      },
      "spatial_distribution": {
        "mean_nearest_neighbor_distance_meters": 13.2,
        "nearest_neighbor_distance_std_meters": 9.7,
        "r_statistic": 0.978,
        "pattern_interpretation": "random"
      }
    }
  }
}

🔍 Analysis Results

Location Characteristics	Value	Interpretation
🏙️ POI Density	1,371 per km²	Ultra-dense location (rural areas: <10)
🍽️ Food Scene	324 establishments	Dining powerhouse in 500m radius
🚲 Transit Access	86 bike facilities	Sustainable transport infrastructure
🏛️ Entertainment	12 theaters + 38 venues	Major entertainment district
🏪 Financial Services	24 banks + 4 ATMs	Active commercial hub

Network Intelligence	Value	Interpretation
🚶 Walkability	5.95 connections/node	Very high pedestrian connectivity
🗺️ Street Pattern	2.056 bearing entropy	Organized grid-like layout
🛣️ Network Density	101.9 km/km²	Dense street network

Spatial Intelligence	Value	Use Case
📊 Shannon Diversity	2.245	High variety → Rich ML features
📈 Simpson Diversity	0.79	Robust POI mix → Stable predictions
🎯 Clustering Pattern	R = 0.978	Random distribution → Uniform coverage

Perfect for: Price prediction models, accessibility scoring, urban planning analysis

🎯 Key Features

Rich POI Analysis (Points of Interest)

• 40+ categories: restaurants, hospitals, schools, transit, entertainment
• Density metrics: POIs per square kilometer by category
• Diversity indices:
  • Shannon diversity: Measures variety and evenness (higher = more diverse)
  • Simpson diversity: Probability two random POIs are different types
• Spatial patterns: clustered, dispersed, or random POI distributions

Street Network Insights

• Connectivity: average connections per intersection
• Total length: meters of streets within radius
• Segment patterns: distribution of street segment lengths
• Bearing analysis: street orientation entropy and grid patterns

Progress Tracking

Mode	Code	Use Case
Standard	show_progress=True, progress_detail='normal'	General use with progress bars
Verbose	show_progress=True, progress_detail='verbose'	Detailed debugging information
Silent	show_progress=False	Batch processing, production

# Example: Verbose progress tracking
features = features_from_location(location, show_progress=True, progress_detail='verbose')

🔬 Scientific Applications

Geospatial Research:

# Compare neighborhood walkability
locations = [
    {'latitude': 40.7580, 'longitude': -73.9855, 'radius_meters': 800},  # Times Square
    {'latitude': 40.7829, 'longitude': -73.9654, 'radius_meters': 800}   # Central Park
]

for loc in locations:
    features = features_from_location(loc)
    walkability_score = (
        features['poi_metrics']['density_metrics']['points_of_interest_per_sqkm'] * 0.4 +
        features['network_metrics']['connectivity_metrics']['average_connections_per_node']['value'] * 100 * 0.6
    )
    print(f"Walkability score: {walkability_score:.1f}")

ML Feature Engineering:

# Generate features for price prediction model
import pandas as pd

properties = pd.read_csv('real_estate.csv')  # lat, lon, price columns
features_list = []

for _, row in properties.iterrows():
    location_features = features_from_location({
        'latitude': row['lat'],
        'longitude': row['lon'], 
        'radius_meters': 1000
    }, show_progress=False)
    
    # Extract key features for ML
    features_list.append({
        'restaurant_density': location_features['poi_metrics']['density_metrics']['restaurant_places_per_sqkm'],
        'transit_access': location_features['poi_metrics']['absolute_counts']['counts_by_category'].get('total_bus_station_places', {}).get('count', 0),
        'street_connectivity': location_features['network_metrics']['connectivity_metrics']['average_connections_per_node']['value'],
        'location_diversity': location_features['poi_metrics']['distribution_metrics']['diversity_metrics']['shannon_diversity_index']
    })

# Add to your ML pipeline
features_df = pd.DataFrame(features_list)
properties = pd.concat([properties, features_df], axis=1)

🛠 Advanced Usage

Batch Processing

# Process multiple locations efficiently
locations = [
    {'latitude': 40.7580, 'longitude': -73.9855, 'radius_meters': 500},
    {'latitude': 40.7829, 'longitude': -73.9654, 'radius_meters': 500},
    {'latitude': 40.7527, 'longitude': -73.9772, 'radius_meters': 500}
]

results = features_from_location(locations, show_progress=True)

Command Line Interface

# Single location analysis
geofeaturekit analyze 40.7580 -73.9855 --radius 500 --verbose

# Batch analysis from file
geofeaturekit batch-analyze locations.json --radius 1000 --output results/

Custom Radius Analysis

# Compare different scales
radii = [200, 500, 1000, 2000]  # meters

for radius in radii:
    features = features_from_location({
        'latitude': 40.7580,
        'longitude': -73.9855, 
        'radius_meters': radius
    })
    
    poi_count = features['poi_metrics']['absolute_counts']['total_points_of_interest']
    print(f"{radius}m radius: {poi_count} POIs")

📖 Key Terms

Term	Definition	Scale
POI	Points of Interest (restaurants, hospitals, schools, ATMs)	Count
Shannon Diversity	Measures variety and evenness of POI types	0-4+ (higher = more diverse)
Simpson Diversity	Probability two random POIs are different types	0-1 (higher = more diverse)
Bearing Entropy	Street grid organization measure	0-4+ (lower = more organized)
R-statistic	Spatial clustering pattern	0-2.1 (<1 clustered, ~1 random, >1 dispersed)
Connectivity	Average connections per street intersection	2-8+ (higher = more walkable)

📊 Output Structure

GeoFeatureKit returns a comprehensive dictionary with four main sections:

{
    'network_metrics': {
        'basic_metrics': {...},      # Node/edge counts, total length
        'density_metrics': {...},    # Per-km² measurements  
        'connectivity_metrics': {...}, # Connection patterns
        'street_pattern_metrics': {...} # Orientation, segment analysis
    },
    'poi_metrics': {
        'absolute_counts': {...},    # Raw POI counts by category
        'density_metrics': {...},    # POIs per km² by category
        'distribution_metrics': {...} # Diversity and spatial patterns
    },
    'units': {
        'area': 'square_meters',
        'length': 'meters', 
        'density': 'per_square_kilometer'
    }
}

🌍 Standards & Quality

• SI Units: All measurements in meters, square kilometers
• Confidence Intervals: Statistical uncertainty for network metrics
• Reproducible: Deterministic results with caching
• Validated: Comprehensive test suite with property-based testing

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

📄 License

MIT License - see LICENSE file for details.

🙏 Acknowledgments

Built with OSMnx, NetworkX, and GeoPandas. Data from OpenStreetMap contributors.

📚 Citation

If you use GeoFeatureKit in your research, please cite:

@software{geofeaturekit2025,
    title={GeoFeatureKit: Geospatial Feature Extraction and Analysis},
    author={Alexander Li},
    year={2025},
    url={https://github.com/lihangalex/geofeaturekit}
}

---

Ready to analyze any location? Start with pip install geofeaturekit and explore geospatial patterns like never before! 🌍